| 00:00.02 | VoodooZ_home | My slug won't recognise my bt dongle at all! |
| 00:00.09 | [g2] | are you feeling less lonely now that I'm here :) |
| 00:00.18 | VoodooZ_home | :) |
| 00:00.18 | [g2] | which one ? |
| 00:00.19 | VoodooZ_home | yeah! |
| 00:00.41 | VoodooZ_home | I got a cheap bluegear twin pack from Radio-scrap |
| 00:01.09 | VoodooZ_home | My RH9 box at work at least gave me some sign the device was inserted. here I get nothing! |
| 00:01.14 | [g2] | IOgear has a Class 1 100 foot radius that Linux recognized |
| 00:01.39 | VoodooZ_home | iogear? |
| 00:01.59 | [g2] | oops, brings up minicom first |
| 00:02.43 | [g2] | # hub.c: new USB device 00:01.0-1, assigned address 2 |
| 00:02.43 | [g2] | usb.c: USB device 2 (vend/prod 0xa12/0x1) is not claimed by any active driver. |
| 00:02.52 | [g2] | needs the driver |
| 00:02.56 | VoodooZ_home | Mine is called bluegear from a company called amazingtech. model # is B091H1. The bluez page says it's supported. |
| 00:03.15 | [g2] | do you get the usb message ? |
| 00:03.25 | VoodooZ_home | Unless I need to enable some extended LUN type of thing? |
| 00:03.30 | VoodooZ_home | Nope. that's the scary part! |
| 00:03.37 | VoodooZ_home | let me reboot it. |
| 00:03.53 | [g2] | bummer kas11 is gone :( |
| 00:04.12 | kas11 | I be here |
| 00:04.18 | VoodooZ_home | hehehe |
| 00:04.31 | VoodooZ_home | Like jbot, always here when in need! |
| 00:04.38 | VoodooZ_home | Get a room! :) |
| 00:04.47 | kas11 | [g2] ;) |
| 00:05.06 | CodeWhacker | VoodooZ_home: ok, I reflashed to openslug and tried my AmbiCom BT dongle. not recognized. |
| 00:05.09 | [g2] | kas11 I'm thinking openslug on a 2.4 kernel in oe, then roll to 2.6 |
| 00:05.30 | [g2] | or do you think we should go straight 2.6 ? |
| 00:05.33 | [g2] | in oe |
| 00:05.46 | CodeWhacker | checked it on the desktop, and it's USB full speed (i.e. 1.1) |
| 00:05.51 | VoodooZ_home | 2.4? why bother? |
| 00:06.16 | [g2] | VoodooZ_home, to setup a jffs2 fs and get oe all straightened out |
| 00:06.20 | VoodooZ_home | Did you check if you BT thing is in the supported list on bluez.org? Most of them are. But some need a F/W upgrade. |
| 00:06.21 | CodeWhacker | so I checked with a 1.1 flash disk. also not recognized. leads me to guess the slug doesn't see anything 1.1 |
| 00:06.25 | VoodooZ_home | I see. |
| 00:06.42 | kas11 | I think ya'll outta forget 2.4...but then again, I am unusual |
| 00:06.42 | [g2] | VoodooZ_home, I run it on the desktop and it talks to my phone |
| 00:07.02 | VoodooZ_home | yeah, I like the fact you can attach a virtual com port to it too! |
| 00:07.03 | CodeWhacker | VoodooZ_home: I have 2 of these. one is plugged into the back panel USB all the time |
| 00:07.16 | CodeWhacker | so I know it works. I use it a lot. |
| 00:07.36 | VoodooZ_home | yeah, I just realised that you were testing 1.1 compatibility. |
| 00:08.00 | VoodooZ_home | Unless you need to load a backward compatibility module for 1.1??? |
| 00:08.14 | kas11 | [g2] did ya ever find the patches directory? theoretically it has been cleaned up |
| 00:08.46 | [g2] | kas11, found it that night, however, the build blew-up with a buildroot toolchain issue |
| 00:08.49 | kas11 | I removed a bunch of my cruft...but it still showed up in the web view...so I dunno |
| 00:09.03 | VoodooZ_home | I think it's delayed. |
| 00:09.14 | [g2] | no that wasn't the problem at all |
| 00:09.32 | VoodooZ_home | I'm talking about the CVS here. |
| 00:09.32 | [g2] | a cvs up doesn't pull new directories |
| 00:09.48 | [g2] | understood, I was talking cvs |
| 00:09.56 | VoodooZ_home | ok. |
| 00:10.12 | kas11 | since I haven't gotten around to fixing the buildroot configs, I have to let the makefile quit and then config twice...then start the main makefile again to finish |
| 00:11.18 | [g2] | kas11 we (I) should just fix the buildroot makefile in cvs |
| 00:11.36 | [g2] | and adjust the patches |
| 00:11.39 | kas11 | anyhow, if it doesn't bult smoothly other than the break for busybox config, I'd like to know so I can fix it |
| 00:11.57 | CodeWhacker | [g2]: concur. I tried to build anyu2 from a fresh cvs co yesterday and didn't have much fun |
| 00:12.20 | [g2] | my original issue was there was no 2.6.9 patch for buildroot |
| 00:12.52 | [g2] | I worked on creating a patch, and got most of the way there, but I think erik/mjn3 fixed it in the mean time |
| 00:12.59 | kas11 | well, I hadn't decided exactly what I wanted in my rootfx so I let it hang...I will get around to expanding my initrd "one day real soon now" (R) |
| 00:13.26 | peteru | [g2]: you need to say cvs up -PAd to pull new dirs |
| 00:13.36 | [g2] | peteru, thx |
| 00:13.55 | kas11 | I need to dump my daemons in with bb to get them to be built automagically too |
| 00:14.12 | CodeWhacker | think I got as far as the busybox break. is tyat patch around somewhere? |
| 00:14.43 | [g2] | kas11, are you building busybox with the 2.6.8 headers ? |
| 00:15.45 | VoodooZ_home | [g2]: Confirmed. My USB1.1 mouse also didn't get acknowledged. Anybody else has any ideas on USB1.1 support?? |
| 00:15.57 | kas11 | with the anyu2 cvs stuff, just config BB in buildroot...and then restart make back in the anyu2 dir and it should pick up where it left off |
| 00:16.15 | kas11 | yes on the 2.6.8 hdrs |
| 00:16.17 | VoodooZ_home | Can someone confirm that USB1.1 devices works on unslung or stock linksys F/W??? |
| 00:17.14 | CodeWhacker | hang on... I'll reflash and test |
| 00:17.22 | VoodooZ_home | Thanks mate. |
| 00:17.29 | Tiersten | Yes. It works fine |
| 00:17.35 | [g2] | VoodooZ_home, I don't think my BT dev is 2.0 and that message was from unslung-2.7alpha |
| 00:17.49 | VoodooZ_home | I believe it's a buggy driver as even the USB2 devices stop being recognise after that. |
| 00:19.08 | VoodooZ_home | I just read today that the bluez team fix 3 bugs in 2.6.9 but I doubt it's related. |
| 00:19.10 | Tiersten | Probably if that happens |
| 00:19.29 | *** join/#nslu2-linux Kaijuu (~Kaijuu@ottawa-hs-209-217-123-11.d-ip.magma.ca) |
| 00:19.50 | VoodooZ_home | wow. I have no clue where to go from here. lucky me. |
| 00:19.52 | Kaijuu | Hello. |
| 00:20.23 | Tiersten | Hi |
| 00:21.25 | kergoth | anyu2 cvs? for what? |
| 00:21.43 | [g2] | kergoth, for kas11 2.6.9 kernel |
| 00:21.49 | [g2] | kas11's |
| 00:22.34 | [g2] | kas11 has been working on anyu2 cvs and not many of his changes are rolled into oe |
| 00:23.00 | kergoth | hand me his patch, it'll take about 30 seconds to add to oe |
| 00:23.49 | kergoth | [g2]: oh, i dont think i told you. i'm going to be working with siti on portage-c. |
| 00:23.52 | [g2] | if [g2] was a girl I'd wanna have kergoth's baby |
| 00:24.05 | [g2] | make that 2 babies |
| 00:24.14 | [g2] | :) cool |
| 00:24.20 | kergoth | started writing an oe metadata handling common module for it |
| 00:24.26 | kergoth | not sure how much we'll be able to share, but there's certainly overlap |
| 00:24.30 | kergoth | i expect we'll figure it out as we go |
| 00:24.49 | [g2] | excellent |
| 00:25.12 | [g2] | As I've mentioned before the metadata is the key to the whole process |
| 00:25.22 | kergoth | course you realize that having a .oe parser in portage would mean that without too much effort, you could use a .oe file to install a package into your gentoo system |
| 00:25.27 | kergoth | which would just be twisted :) |
| 00:25.43 | [g2] | it's really just data to me |
| 00:25.48 | kergoth | yep |
| 00:25.56 | [g2] | .oe, .ebuild, .deb, .ipkg |
| 00:26.01 | kergoth | just a matter of how you manipulate it. portage is limited because it relies on shell |
| 00:26.07 | kergoth | less control over the data |
| 00:26.33 | CodeWhacker | VoodooZ_home: unslung 1.15 recognizes my bt dongle (but no driver present) |
| 00:26.41 | [g2] | I think kergoth takes the road less traveled, but the proper one in many cases |
| 00:26.57 | VoodooZ_home | crap. It must specific to openslug then. Unless linksys hacked stuff in??? |
| 00:27.07 | peteru | kergoth: any way of speeding up or skipping the initial parts of an oemake? the parsing takes a long time and half the time I am just trying to rebuild the same package after a few patches are applied. |
| 00:27.09 | [g2] | CodeWhacker, which BT dongle do you have ? |
| 00:27.12 | Tiersten | ah yes. can you get rid of that skbuff hack before sending any patches to OE? |
| 00:27.13 | kergoth | heh, i may not be the best of coders, but i do a fairly good job of taking the clean, maintainable approaches. |
| 00:27.19 | CodeWhacker | [g2]: AmbiCom |
| 00:27.19 | kergoth | peteru: it already caches. |
| 00:27.35 | [g2] | peteru: are you running Psyco and using the packages hack ? |
| 00:27.37 | peteru | I know, but it still takes a couple of minutes |
| 00:27.40 | kergoth | peteru: you can play with the copy-on-write implementation patches |
| 00:27.48 | kergoth | peteru: which cuts the parse time by about a fifth or so |
| 00:27.55 | peteru | even with psyco |
| 00:27.59 | kergoth | i have a new metadata implementation that cuts it by about a third, but it isnt ready for prime time |
| 00:28.04 | Tiersten | reverting a change that the main kernel tree has made to get around patching the CSR driver isn't a good idea |
| 00:28.42 | [g2] | Tiersten, agreed |
| 00:29.07 | [g2] | I'd forgo the Intel CSR on the 2.6.9 for now and use my usbnet dongle |
| 00:29.19 | Tiersten | Actually, we're patching ixp425_eth.c anyway. Why do we still have the skbuff change still? |
| 00:29.32 | [g2] | but I'd love to be at the spot where that was the current issue |
| 00:30.08 | Tiersten | CSR is really turning out to be a PITA |
| 00:31.02 | peteru | kergoth: where would I look for the COW patches? |
| 00:31.06 | Tiersten | [g2]: heh |
| 00:31.17 | kergoth | peteru: oe mailing list. |
| 00:31.32 | Tiersten | You'd still need that microcode you load into the NPEs anyway |
| 00:31.42 | [g2] | kas11 can you generate a 2.6.9 patch fro kergoth and all of us ? |
| 00:31.47 | Tiersten | That won't be trivial to reverse engineer considering we don't even know what it's doing |
| 00:32.18 | peteru | trying to find licensing loopholes doesn't pay off in the long run. |
| 00:32.52 | peteru | campainging for corporate attitude change can often be more productive :( |
| 00:33.28 | kergoth | licensing loopholes? the limitation is you cant distribute the source. it really isnt that big a deal |
| 00:33.34 | kergoth | annoying, but not really problematic |
| 00:33.41 | kas11 | [g2] there is a single patch in the anyu2 cvs directory now...but it does revert skbuff.h |
| 00:33.43 | CodeWhacker | [g2]: is |
| 00:33.55 | CodeWhacker | [g2]: is .config.bb around somewhere? |
| 00:35.05 | kas11 | the skbuffs are private to the CSR so I cannot for the life of me see the harm...but if someone wants to screw with it, they outta have at it |
| 00:35.59 | [g2] | CodeWhacker, dunno (I'm sure it's somewhere) for the anyu2 ? |
| 00:36.18 | CodeWhacker | [g2]: yep. that's where my build breaks right now |
| 00:36.42 | Tiersten | The point is that the newer kernels are trying to separate the protocol specific stuff e.g. ethernet |
| 00:36.50 | kas11 | the whole damn csr is overly complicated and a major overkill for the nslu2...and since the PHY isn't even supported...the whole thing is a pita |
| 00:36.59 | Tiersten | You just have to use the right accessor function |
| 00:37.12 | Tiersten | eth_hdr() from what I remember |
| 00:37.59 | Tiersten | You've sort of got a mistake in nslu2-pci.c |
| 00:38.14 | Tiersten | You're still configuring INTD even though it's been reused as the power down output |
| 00:38.23 | [g2] | with 10**6 all bugs are shallow |
| 00:38.27 | [g2] | eyes |
| 00:38.32 | Tiersten | *shrug* Linksys do it but it's a bit of a hack |
| 00:39.22 | [g2] | kas11 does it recognize the flash partitions correctly ? |
| 00:39.56 | kas11 | haven't even looked at nslu2-pci...mrakes did that |
| 00:40.19 | Tiersten | ah. you're listed as the last person but it seems you just moved it |
| 00:40.31 | [g2] | do you have a dmesg of the boot ? |
| 00:40.43 | [g2] | could you pls post it on the wiki ? |
| 00:40.43 | Tiersten | Remove INTD from it if you've got the time. No point having it configuring it as a interrupt when it's not even used as an input |
| 00:41.09 | kas11 | I'm running out of ram...no sense if messing with flash when I am changing things so often |
| 00:42.16 | [g2] | here's my $0.02 on openslug |
| 00:42.36 | [g2] | I'd like to see a stable 2.6.9 kernel |
| 00:42.52 | [g2] | I'd like to see a very tiny ext2 ramdisk |
| 00:43.01 | [g2] | I'd like to see a pretty big jffs2 part |
| 00:43.48 | Tiersten | I'd like a pony! ;) |
| 00:43.59 | CodeWhacker | [g2]: with a jffs2 fs, why do we need a ramdisk? |
| 00:44.01 | [g2] | LOL |
| 00:44.17 | Tiersten | Doing JFFS2 needs RedBoot to be replaced though unless you're okay with losing 256K of flash |
| 00:44.22 | [g2] | CodeWhacker, it's a stopping point on the way to squshfs |
| 00:44.41 | kergoth | you dont need to use ramdisk for that. |
| 00:44.45 | [g2] | I'm plenty OK with losing 256K of flash |
| 00:44.52 | VoodooZ_home | here we go. we've been full circle again.. |
| 00:44.56 | kergoth | could do the same thing with tmpfs. |
| 00:45.21 | [g2] | ok skip the JFFS2 part |
| 00:45.31 | [g2] | a small ramdisk that boot to the hd |
| 00:46.02 | [g2] | I was within the scum on my teeth of getting the 2.4 kernel booting directly to the hd |
| 00:46.36 | [g2] | There was 1 /mount -t proc cmd that was the only thing in the way |
| 00:47.51 | [g2] | Tiersten, this is the line you'd like to go away ? + IRQ_NSLU2_PCI_INTD |
| 00:48.10 | Tiersten | There's a whole bunch in nslu2-pci.c |
| 00:48.19 | Tiersten | It's not a major thing. Just cleaner |
| 00:48.38 | Tiersten | The stock kernel from Linksys actually does the same thing but has it set as an output to do the power down stuff |
| 00:49.27 | [g2] | Tiersten, I don't see any skbuff changes in the nslu2_2.6.9-rc2.patch file from cvs |
| 00:49.45 | kas11 | As I recall, nslu2-pci.c is a clone of what DSaxena was doing with ixp425-pci.c |
| 00:49.58 | Tiersten | rc2? |
| 00:50.10 | Tiersten | I'm looking at the one here - http://cvs.sourceforge.net/viewcvs.py/nslu/anyu2/patches/ |
| 00:50.25 | kas11 | nslu2_2.6.9-rc2.patch isn't the patch I am using |
| 00:50.44 | [g2] | ah.. the patches/. |
| 00:51.26 | [g2] | CodeWhacker, you're anyu2 build get here too ? cp: cannot stat `/home/tom/dev/nslu2/anyu2/patches/.config.bb': No such file or directory |
| 00:51.42 | CodeWhacker | that's exactly where it dies |
| 00:51.46 | kas11 | [g2] my patch may have changed more recently than when you downloaded it |
| 00:52.15 | [g2] | kas11 I cvs up -PAd 10-15 minutes ago |
| 00:52.20 | kas11 | k |
| 00:52.29 | [g2] | 64d6f6650f29dd25f9735374ddb79a2e nslu2_2.6.9.patch |
| 00:52.55 | kas11 | that be it |
| 00:53.21 | CodeWhacker | eh? |
| 00:53.30 | CodeWhacker | 3fe71fb2eb0af9f1b0aed3fbced5d086 patches/nslu2_2.6.9.patch |
| 00:53.51 | CodeWhacker | and I updated 30 seconds ago |
| 00:54.12 | CodeWhacker | am I missing something? |
| 00:54.28 | [g2] | you're pulling ext not pserver right ? |
| 00:54.36 | CodeWhacker | (besides my sanity, which left long ago) |
| 00:54.42 | [g2] | :) |
| 00:55.26 | CodeWhacker | yep, ext |